Understanding the mechanics of creep deformation to develop a surrogate model for contact assessment in CANDU (R) fuel channels

作者：Pandey M D^*; Tallavo F J; Christodoulou N C^*; Leitch B; Bickel G A

来源：Nuclear Engineering and Design, 2018, 330: 141-156.

DOI：10.1016/j.nucengdes.2018.01.032

• Fuel channel
• Creep deformation
• Finite element method
• Surrogate model
• Pressure tube
• Calandria tube
• Contact assessment
• Zirconium alloy
• Probabilistic assessments

A key element of the fuel channel life cycle management in CANDU (R) reactors is to prevent contact between the pressure tube (PT) and the calandria tube (CT) and to avoid the development of hydride blisters that lead to delayed hydride cracking of the PT. The PT-CT contact is the result of in-reactor deformation due to irradiation induced creep of the fuel channel assembly, which in turn is affected by uncertainties associated with various parameters like material properties, dimensional changes in the channel and boundary conditions (e.g., end slopes) of the channel. To account for these uncertainties, probabilistic assessment methods are developed to evaluate the risk of PT-CT contact and demonstrate compliance with provisions of the CSA Standard N285.8. Currently, a simulation is based on probabilistic assessments in which input parameters to a finite element model (FEM) of creep deformation are randomly sampled from their respective distributions. A simulation model involves numerous repetitive solutions of the FEM model to determine the probability distribution of PT-CT gap and the time to contact. Since the creep deformation analysis using FEM is computationally involved, this brute force Monte Carlo simulation method is not an efficient way to carry out the probabilistic assessment of the reactor core. This paper proposes a new line of thinking for probabilistic assessments of PT-CT contact in fuel channels, which is based on replacing a full FEM model by a surrogate model of a much simpler analytical form. The surrogate model not only simplifies the creep deformation analysis, but also provides a more logical basis for probabilistic assessments. This paper presents an insightful analysis of creep deformation and shows that a simple surrogate model can be developed to predict the PT-CT gap as a linear function of two primary random variables, namely, a creep factor and end slopes. This simplified representation has a far reaching effect on the probabilistic assessment of fuel channels.

• 出版日期2018-4-15